Human IL-8 is a cytokine which has variously been called neutrophil-activating protein, neutrophil chemotactic factor (NCF) and T-cell chemotactic factor. IL-8 can be secreted by several types of cells upon appropriate stimulation. IL-8 is secretedby activated monocytes and macrophages as well as by embryonic fibroblasts.
IL-8 is known to induce neutrophil migration and to activate functions of neutrophils such as degranulation, release of superoxide anion and adhesion to the endothelial cell monolayer. There are a number of conditions that are known to involve leukocyte infiltration into lesions. These include pulmonary diseases such as pulmonary cystic fibrosis, idiopathic pulmonary fibrosis, adult respiratory distress syndrome, sarcoidosis and empyema; dermal diseases such as psoriasis, rheumatoid arthritis, Crohn's Disease; and in inflammatory bowel disease (McElvaney, N. G. et al. J Clin Invest (1992) 90:1296-1301; Lynch III, J. P. et al. Am Rev Respir Dis (1992) 145:1433-1439; Donnelly, S. C. et al. Lancet (1993) 341:643-647; Car, B. D. et al. Am J Respir Crit Care Med (1994) 149:655-659; Antony, V. B. et al. J Immunol (1993) 151:7216-7223; Takematsu, H. et al. Arch Dermatol (1993) 129:74-80; Brennan, F. M. et al. Eur J Immunol (1990) 20:2141-2144; Izzo, R. S. et al. Scand J Gastroenterol (1993) 28:296-300; Izzo, R. S. et al. Am J Gastroenterol (1992) 87:1447-1452).
The amino acid sequence characterizing human IL-8 was described by Matsushima, et al. in PCT application WO89/08665. More recently, Yoshimura, T., et al. in Mol Immunol (1989) 26:87-93 showed that monocyte-derived IL-8 was evidently variably processed at the N-terminus and that the IL-8 originally disclosed by Matsushima et al. was accompanied by two forms of the factor which had seven or five additional amino acids at the N-terminus. The longest form accounted for about 8%, the next longest form for about 47%, and the shortest form for about 45% of the total IL-8 derived from monocytes.
WO89/08665 also reports the production of murine monoclonal antibodies immunoreactive with this protein (called neutrophil chemotactic factor in this publication). An additional murine monoclonal antibody immunospecific for IL-8 and designated WS-4 was prepared and reported by Ko, Y. C. et al. J Immunol Meth (1992) 149:227-235. DNA encoding heavy and light chains of this antibody was recovered from the hybridoma which produced it, as described below. Additional murine-derived antibodies to human IL-8 were reported by Boylan, A. M. et al. J Clin Invest (1992) 89:1257-1267 (A5.12.14); in PCT application WO92/04372 (Anti-Pep1 AND Anti-Pep3) and by Mulligan, M. S. et al. (J Immunol (1993) 150:5585-5595) (DM/C7).
WS-4 has been shown to inhibit the binding of rabbit IL-8 to rabbit neutrophils by Harada, A. et al. International Immunol (1993) 5:681-690. Administration of WS-4 to rabbits also reduced ischemia/reperfusion injury in the lung (Sekido, N. et al. Nature (1993) 365:654-657); reduced LPS-induced dermatitis (Harada et al. (supra)); and ameliorated LPS or IL-1 induced arthritis (Akahoshi, T. et al. Lymphokine and Cytokine Res (1994) 13:113-116). Administration of DM/C7 intratracheally to rats was protective with respect to inflammatory lung injury (Mulligan, M. S., et al. (supra)). Thus, animal models have demonstrated that antibodies to IL-8 are effective in treating diseases and conditions which are characterized by inflammation.
It would therefore be useful to have antibodies immunoreactive with human IL-8 which would be compatible with the human immune system so that these antibodies could be used as therapeutic agents. It is known that murine antibodies are highly immunogenic in humans, thus limiting their value as therapeutic agents. Furthermore, murine antibodies have low circulating half-lives in humans. It would be ideal to prepare anti-IL-8 antibodies which do not raise antibodies against themselves in human patients and which retain their effectiveness against human IL-8.
Other nonhuman antibodies have been "humanized" with varying degrees of success in the past. In a straightforward and simple but incomplete approach, chimeric antibodies are prepared, generally using recombinant techniques, which contain nonhuman variable regions and human constant regions. This eliminates the constant region as an immunogen in human patients, but the possibility of an immune response to the foreign variable region remains (LoBuglio, A. F. et al. Proc Natl Acad Sci USA (1989) 86:4220-4224).
A more sophisticated approach focuses not only on providing human-derived constant regions, but modifying the variable regions as well so as to reshape them as closely as possible to human form. It is known that the variable regions of both heavy and light chains contain three complementarity-determining regions (CDRs) which vary in response to the antigens in question and determine binding capability, flanked by four framework regions (FRs) which are relatively conserved in a given species and which putatively provide a scaffolding for the CDRs. When nonhuman antibodies are prepared with respect to a particular antigen, the variable regions can be "reshaped" or "humanized" by grafting CDRs derived from nonhuman antibody on the FRs present in the human antibody to be modified. Application of this approach to various antibodies has been reported by Sato, K., et al. Cancer Res (1993) 53:851-856. Riechmann, L., et al. Nature (1988) 332:323-327; Verhoeyen, M., et al. Science (1988) 239:1534-1536; Kettleborough, C. A., et al. Protein Engineering (1991) 4:773-3783; Maeda, H., et al. Human Antibodies Hybridoma (1991) 2:124-134; Gorman, S. D., et al. Proc Natl Acad Sci USA (1991) 88:4181-4185; Tempest, P. R., et al. Bio/Technology (1991) 9:266-271; Co, M. S., et al., Proc Natl Acad Sci USA (1991) 88:2869-2873; Carter, P., et al., Proc Natl Acad Sci USA (1992) 89:4285-4289; and Co, M. S. et al. J Immunol (1992) 148:1149-1154.
In general, these techniques involve identifying the regions responsible for binding and those responsible for supporting these binding domains and then grafting these CDRs onto human framework regions which support these binding regions. It will be evident that it is impossible simply to extrapolate this approach from one antibody raised against one antigen to the next. Antibodies with respect to a particular antigen will require carefully designed and particularly devised modifications to obtain the desired result.